
import pandas as pd
import numpy as np
import statsmodels.api as sm
import matplotlib.pyplot as plt
from matplotlib import style
style.use('ggplot') or plt.style.use('ggplot')
plt.rcParams["figure.figsize"] = (12,10)

pd.set_option('display.max_rows', None, 'display.max_columns', None)

elec = pd.read_excel('DatosEleccionesEspaña.xlsx')
elec.head()


elec.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8119 entries, 0 to 8118
Data columns (total 41 columns):
 #   Column                        Non-Null Count  Dtype  
---  ------                        --------------  -----  
 0   Name                          8119 non-null   object 
 1   CodigoProvincia               8119 non-null   int64  
 2   CCAA                          8119 non-null   object 
 3   Population                    8119 non-null   int64  
 4   TotalCensus                   8119 non-null   int64  
 5   AbstentionPtge                8119 non-null   float64
 6   AbstencionAlta                8119 non-null   int64  
 7   Izda_Pct                      8119 non-null   float64
 8   Dcha_Pct                      8119 non-null   float64
 9   Otros_Pct                     8119 non-null   float64
 10  Izquierda                     8119 non-null   int64  
 11  Derecha                       8119 non-null   int64  
 12  Age_0-4_Ptge                  8119 non-null   float64
 13  Age_under19_Ptge              8119 non-null   float64
 14  Age_19_65_pct                 8119 non-null   float64
 15  Age_over65_pct                8119 non-null   float64
 16  WomanPopulationPtge           8119 non-null   float64
 17  ForeignersPtge                8119 non-null   float64
 18  SameComAutonPtge              8119 non-null   float64
 19  SameComAutonDiffProvPtge      8119 non-null   float64
 20  DifComAutonPtge               8119 non-null   float64
 21  UnemployLess25_Ptge           8119 non-null   float64
 22  Unemploy25_40_Ptge            8119 non-null   float64
 23  UnemployMore40_Ptge           8119 non-null   float64
 24  AgricultureUnemploymentPtge   8119 non-null   float64
 25  IndustryUnemploymentPtge      8119 non-null   float64
 26  ConstructionUnemploymentPtge  8119 non-null   float64
 27  ServicesUnemploymentPtge      8119 non-null   float64
 28  totalEmpresas                 8114 non-null   float64
 29  Industria                     7931 non-null   float64
 30  Construccion                  7980 non-null   float64
 31  ComercTTEHosteleria           8110 non-null   float64
 32  Servicios                     8057 non-null   float64
 33  ActividadPpal                 8119 non-null   object 
 34  inmuebles                     7981 non-null   float64
 35  Pob2010                       8112 non-null   float64
 36  SUPERFICIE                    8110 non-null   float64
 37  Densidad                      8119 non-null   object 
 38  PobChange_pct                 8112 non-null   float64
 39  PersonasInmueble              7981 non-null   float64
 40  Explotaciones                 8119 non-null   int64  
dtypes: float64(30), int64(7), object(4)
memory usage: 2.5+ MB


elec.nunique()

Name                            8102
CodigoProvincia                   52
CCAA                              19
Population                      3597
TotalCensus                     3310
AbstentionPtge                  5675
AbstencionAlta                     2
Izda_Pct                        6569
Dcha_Pct                        6682
Otros_Pct                       4319
Izquierda                          2
Derecha                            2
Age_0-4_Ptge                    3761
Age_under19_Ptge                5891
Age_19_65_pct                   6215
Age_over65_pct                  6778
WomanPopulationPtge             4524
ForeignersPtge                  2329
SameComAutonPtge                6151
SameComAutonDiffProvPtge        4207
DifComAutonPtge                 5574
UnemployLess25_Ptge             2342
Unemploy25_40_Ptge              2681
UnemployMore40_Ptge             2751
AgricultureUnemploymentPtge     2525
IndustryUnemploymentPtge        2538
ConstructionUnemploymentPtge    2505
ServicesUnemploymentPtge        2904
totalEmpresas                   1225
Industria                        307
Construccion                     456
ComercTTEHosteleria              802
Servicios                        757
ActividadPpal                      5
inmuebles                       3087
Pob2010                         3624
SUPERFICIE                      8109
Densidad                           4
PobChange_pct                   3048
PersonasInmueble                 282
Explotaciones                    758
dtype: int64


factors = list(elec.loc[:, elec.nunique()<= 10])
elec[factors] = elec[factors].astype('category')
elec['CodigoProvincia'] = elec['CodigoProvincia'].astype('object')


#Vamos a ver ahora un poco más de información de las variables numéricas para luego elegir 10 de ellas

elec.describe()


elec1 = elec[["CCAA","Population", "AbstentionPtge", "Izda_Pct", "Age_over65_pct", "WomanPopulationPtge", "UnemployLess25_Ptge", 
             "UnemployMore40_Ptge", "SUPERFICIE", "totalEmpresas", "Pob2010"]]
elec1.set_index(['CCAA'], inplace = True)
elec1.head()


elec1.info()

<class 'pandas.core.frame.DataFrame'>
Index: 8119 entries, Extremadura to CastillaLeón
Data columns (total 10 columns):
 #   Column               Non-Null Count  Dtype  
---  ------               --------------  -----  
 0   Population           8119 non-null   int64  
 1   AbstentionPtge       8119 non-null   float64
 2   Izda_Pct             8119 non-null   float64
 3   Age_over65_pct       8119 non-null   float64
 4   WomanPopulationPtge  8119 non-null   float64
 5   UnemployLess25_Ptge  8119 non-null   float64
 6   UnemployMore40_Ptge  8119 non-null   float64
 7   SUPERFICIE           8110 non-null   float64
 8   totalEmpresas        8114 non-null   float64
 9   Pob2010              8112 non-null   float64
dtypes: float64(9), int64(1)
memory usage: 697.7+ KB


elec1.describe()


# Proporción de missings por observación
elec1['prop_missings'] = elec1.apply(lambda x: x.isna().sum()/x.count()*100,axis=1)

# Cantidad y porcentaje de observaciones con al menos un dato perdido:
qty = len(elec1[elec1['prop_missings']>0])
ptg_miss = round(len(elec1[elec1['prop_missings']>0])/len(elec1['prop_missings'])*100, 2)
print('Cantidad de observaciones con datos perdidos: ', qty, '\nPorcentaje: ', ptg_miss)

Cantidad de observaciones con datos perdidos:  12 
Porcentaje:  0.15

C:\Users\ferli\AppData\Local\Temp\ipykernel_27504\2676930901.py:2: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  elec1['prop_missings'] = elec1.apply(lambda x: x.isna().sum()/x.count()*100,axis=1)


elec1[elec1['prop_missings']>0]


elec1.loc['Melilla']

Population             85584.000000
AbstentionPtge            48.650000
Izda_Pct                  34.417000
Age_over65_pct             9.684000
WomanPopulationPtge       49.064000
UnemployLess25_Ptge       16.363000
UnemployMore40_Ptge       37.472000
SUPERFICIE                      NaN
totalEmpresas           4349.000000
Pob2010                76034.000000
prop_missings             11.111111
Name: Melilla, dtype: float64


elec2 = elec1[(elec1['prop_missings']==0) | (elec1.index =='Melilla')].drop('prop_missings', axis = 1)
elec2.info()

<class 'pandas.core.frame.DataFrame'>
Index: 8108 entries, Extremadura to CastillaLeón
Data columns (total 10 columns):
 #   Column               Non-Null Count  Dtype  
---  ------               --------------  -----  
 0   Population           8108 non-null   int64  
 1   AbstentionPtge       8108 non-null   float64
 2   Izda_Pct             8108 non-null   float64
 3   Age_over65_pct       8108 non-null   float64
 4   WomanPopulationPtge  8108 non-null   float64
 5   UnemployLess25_Ptge  8108 non-null   float64
 6   UnemployMore40_Ptge  8108 non-null   float64
 7   SUPERFICIE           8107 non-null   float64
 8   totalEmpresas        8108 non-null   float64
 9   Pob2010              8108 non-null   float64
dtypes: float64(9), int64(1)
memory usage: 696.8+ KB


import sklearn.impute as skl_imp

imputer_knn = skl_imp.KNNImputer(n_neighbors=3)

elec_imp = pd.DataFrame(imputer_knn.fit_transform(elec2),columns=elec2.columns, index = elec2.index)
elec_imp[['Population','SUPERFICIE']].groupby('CCAA').sum()


elec_m = elec_imp.groupby('CCAA').agg({
    'Population':'sum', 'SUPERFICIE':'sum', 'totalEmpresas':'sum', 'Pob2010':'sum', 
    'AbstentionPtge':'mean', 'Izda_Pct':'mean', 'Age_over65_pct':'mean', 'WomanPopulationPtge':'mean', 
    'UnemployLess25_Ptge':'mean', 'UnemployMore40_Ptge': 'mean'})
elec_m


from sklearn.preprocessing import scale

s_elec = scale(elec_m)
pd.DataFrame(s_elec)


from sklearn.cluster import AgglomerativeClustering
from scipy.cluster.hierarchy import dendrogram
from time import time

def plot_dendogram(model, **kwargs):    
  counts = np.zeros(model.children_.shape[0])
  n_samples = len(model.labels_)
  for i, merge in enumerate(model.children_):
    current_count = 0
    for idx in merge:
      if idx < n_samples:
        current_count += 1  # leaf node
      else:
        current_count += counts[idx - n_samples]
    counts[i] = current_count
  
  linkage_matrix = np.column_stack([model.children_, model.distances_,
                                        counts]).astype(float)
  
  # Plot
  dendrogram(linkage_matrix, **kwargs)
  plt.show()


for linkage in ("ward", "average", "complete", "single"):
    clustering = AgglomerativeClustering(linkage=linkage, n_clusters=None,
                            distance_threshold = 0)
    t0 = time()
    clustering.fit(s_elec)
    print("%s :\t%.2fs" % (linkage, time() - t0))
    
    plt.clf()
    plot_dendogram(clustering)

ward :	0.00s

average :	0.00s

complete :	0.00s

single :	0.00s


modelo_hclust_ward = AgglomerativeClustering(
                            affinity = 'euclidean',
                            linkage  = 'ward',
                            n_clusters  = 4
                     )
modelo_hclust_ward.fit(X=s_elec)

cluster_labels_w = modelo_hclust_ward.fit_predict(s_elec)


# Silueta Clustering Ward
from sklearn.metrics import silhouette_score

silhouette_score(s_elec, cluster_labels_w)

0.36896276000054923


#Indice de Calinski-Harabasz

from sklearn import metrics
metrics.calinski_harabasz_score(s_elec, cluster_labels_w)

9.37060337359143


# Veamos los centroides:
from sklearn.neighbors import NearestCentroid

clf = NearestCentroid()
clf.fit(s_elec, cluster_labels_w)
print(clf.centroids_)

[[ 1.45125569  0.3205507   1.46793048  1.45570065 -0.26140586 -0.68508653
  -0.10062442  0.55776809 -0.25890451  0.63335438]
 [-0.37186235  1.0259725  -0.3992826  -0.36341829 -0.99113744  0.11612508
   1.30514275 -1.66520014 -0.62440882  0.18099554]
 [-0.96691934 -0.90258449 -0.94391732 -0.97606197  2.37433231 -0.30513918
  -1.74012564  0.54512157  2.13031548 -2.30312703]
 [-0.4793738  -0.48768431 -0.48183592 -0.48408827  0.0653643   0.44640134
  -0.1546497   0.34771462 -0.05855915  0.0894375 ]]


from scipy.spatial.distance import cdist, pdist
from sklearn.cluster import KMeans

def scree_plot_kmeans(data,n_max):
  range_n_clusters = range(2, n_max)
  X_scaled = scale(data)
  inertias = []
  silhouette = []
  var_perc = []
  
  for n_clusters in range_n_clusters:
      modelo_kmeans = KMeans(
                          n_clusters   = n_clusters, 
                          n_init       = 20, 
                          random_state = 123
                      )
      modelo_kmeans.fit(X_scaled)
      cluster_labels = modelo_kmeans.fit_predict(X_scaled)
      inertias.append(modelo_kmeans.inertia_)
      silhouette.append(silhouette_score(X_scaled, cluster_labels))
      tss = sum(pdist(X_scaled)**2)/X_scaled.shape[0]
      bss = tss - modelo_kmeans.inertia_
      var_perc.append(bss/tss*100)
      
  fig, ax = plt.subplots(1, 3, figsize=(16, 6))
  ax[0].plot(range_n_clusters, inertias, marker='o')
  ax[0].set_title("Scree plot Varianza intra")
  ax[0].set_xlabel('Número clusters')
  ax[0].set_ylabel('Intra-cluster (inertia)')
  
  ax[1].plot(range_n_clusters, silhouette, marker='o')
  ax[1].set_title("Scree plot silhouette")
  ax[1].set_xlabel('Número clusters')
  ax[1].set_ylabel('Media índices silhouette');
  
  ax[2].plot(range_n_clusters, var_perc, marker='o')
  ax[2].set_title("Scree plot % Varianza")
  ax[2].set_xlabel('Número clusters')
  ax[2].set_ylabel('% de varianza explicada')


scree_plot_kmeans(elec_m,10)
plt.show()


modelo_kmeans = KMeans(n_clusters=5, n_init=25, random_state=123)
modelo_kmeans.fit(X=s_elec)

KMeans(n_clusters=5, n_init=25, random_state=123)


print('Varianza intra: ' + str(modelo_kmeans.inertia_))
print('\nCentroides')
print(modelo_kmeans.cluster_centers_)
print('\nEtiquetas')
print(modelo_kmeans.labels_[:5])

Varianza intra: 44.35015944227528

Centroides
[[-0.96691934 -0.90258449 -0.94391732 -0.97606197  2.37433231 -0.30513918
  -1.74012564  0.54512157  2.13031548 -2.30312703]
 [ 1.20786697 -0.11266759  1.38184     1.21889477 -0.27968948 -1.29276209
  -0.0352914   0.53208909 -0.73229776  1.05621747]
 [-0.4793738  -0.48768431 -0.48183592 -0.48408827  0.0653643   0.44640134
  -0.1546497   0.34771462 -0.05855915  0.0894375 ]
 [-0.37186235  1.0259725  -0.3992826  -0.36341829 -0.99113744  0.11612508
   1.30514275 -1.66520014 -0.62440882  0.18099554]
 [ 2.42481057  2.05342382  1.8122924   2.4029242  -0.18827138  1.74561569
  -0.36195653  0.66048413  1.63466851 -1.058098  ]]

Etiquetas
[4 3 2 2 2]


cluster_labels_k = modelo_kmeans.labels_

print('Silueta:', silhouette_score(s_elec, cluster_labels_k))
print('Indice de Calinski-Harabasz:', metrics.calinski_harabasz_score(s_elec, cluster_labels_k))

Silueta: 0.3847878352621361
Indice de Calinski-Harabasz: 11.494309115518341


elec_m['cluster_ward'] = cluster_labels_w
elec_m['cluster_kmeans'] = cluster_labels_k

elec_m[elec_m.cluster_ward != elec_m.cluster_kmeans][['cluster_ward', 'cluster_kmeans']]


elec_m.sort_values(by = 'cluster_kmeans').drop('cluster_ward', axis = 1)


from sklearn import decomposition as dc
from pca import pca

# Initialize pca with default parameters
pcaModel = pca(normalize=True,n_components=2)
results = pcaModel.fit_transform(elec_m.drop('cluster_ward', axis = 1))

[pca] >Processing dataframe..
[pca] >Normalizing input data per feature (zero mean and unit variance)..
[pca] >The PCA reduction is performed on the [11] columns of the input dataframe.
[pca] >Fit using PCA.
[pca] >Compute loadings and PCs.
[pca] >Compute explained variance.
[pca] >Outlier detection using Hotelling T2 test with alpha=[0.05] and n_components=[2]
[pca] >Outlier detection using SPE/DmodX with n_std=[2]


pcaModel.plot()

(<Figure size 1500x1000 with 1 Axes>,
 <AxesSubplot:title={'center':'Cumulative explained variance\n 2 Principal Components explain [65.94%] of the variance.'}, xlabel='Principle Component', ylabel='Percentage explained variance'>)

<Figure size 1200x1000 with 0 Axes>


# Acceso a las cargas
pcaModel.results['loadings']


# Visualización biplot para interpretar componentes
pcaModel.biplot(legend=False)

[pca] >Plot PC1 vs PC2 with loadings.
[colourmap]> Warning: Colormap [Set1] can not create [19] unique colors! Available unique colors: [9].
[colourmap]> Warning: Colormap [Set1] can not create [19] unique colors! Available unique colors: [9].

(<Figure size 1500x1000 with 1 Axes>,
 <AxesSubplot:title={'center':'2 Principal Components explain [65.94%] of the variance'}, xlabel='PC1 (37.6% expl.var)', ylabel='PC2 (28.3% expl.var)'>)


# Acceso a scores
pcaModel.results['PC']

# Unimos variable de cluster_kmeans
elec_pca = pcaModel.results['PC'].join(elec_m['cluster_kmeans'])
elec_pca.sort_values(by = 'cluster_kmeans')


# Visualización del biplot por cluster

import plotly.express as px

fig = px.scatter(elec_pca, x='PC1',y='PC2', color='cluster_kmeans', text=elec_m.index)
fig.show()

	Name	CodigoProvincia	CCAA	Population	TotalCensus	AbstentionPtge	AbstencionAlta	Izda_Pct	Dcha_Pct	Otros_Pct	Izquierda	Derecha	Age_0-4_Ptge	Age_under19_Ptge	Age_19_65_pct	Age_over65_pct	WomanPopulationPtge	ForeignersPtge	SameComAutonPtge	SameComAutonDiffProvPtge	DifComAutonPtge	UnemployLess25_Ptge	Unemploy25_40_Ptge	UnemployMore40_Ptge	AgricultureUnemploymentPtge	IndustryUnemploymentPtge	ConstructionUnemploymentPtge	ServicesUnemploymentPtge	totalEmpresas	ActividadPpal	inmuebles	Pob2010	SUPERFICIE	Densidad	PobChange_pct	PersonasInmueble	Explotaciones
0	Abadía	10	Extremadura	336	282	20.213	0	60.444	35.555	1.778	1	0	3.869	18.155	55.059	26.785	44.048	0.89	79.762	0.298	19.345	2.381	54.762	42.857	4.762	9.524	11.905	73.810	15.0	Otro	216.0	326.0	4507.5593	MuyBaja	3.07	1.56	28
1	Abertura	10	Extremadura	429	364	25.275	0	54.779	44.118	0.368	1	0	1.632	13.055	56.643	30.304	50.117	1.63	90.909	2.797	7.226	16.216	32.432	51.351	8.108	8.108	10.811	67.568	11.0	Otro	382.0	459.0	6270.7646	MuyBaja	-6.54	1.12	67
2	Acebo	10	Extremadura	569	569	27.241	0	44.203	53.140	0.966	0	1	1.230	9.139	54.834	36.028	49.033	0.70	78.910	0.703	18.102	8.197	36.066	55.738	22.951	9.836	13.115	49.180	49.0	Otro	918.0	674.0	5702.1000	MuyBaja	-15.58	0.62	74
3	Acehúche	10	Extremadura	822	704	30.114	1	50.813	45.325	0.000	1	0	4.258	14.964	60.098	24.940	51.095	0.12	93.917	0.487	5.109	7.407	61.111	31.481	16.667	5.556	16.667	59.259	50.0	Otro	599.0	842.0	9106.4649	MuyBaja	-2.38	1.37	66
4	Aceituna	10	Extremadura	623	540	30.185	1	44.562	49.867	0.796	0	1	3.531	15.569	59.391	25.042	48.154	0.64	93.258	0.161	4.173	15.385	48.077	36.538	21.154	0.000	11.538	61.538	22.0	Otro	394.0	625.0	4007.6141	MuyBaja	-0.32	1.58	96

	Population	TotalCensus	AbstentionPtge	Izda_Pct	Dcha_Pct	Otros_Pct	Age_0-4_Ptge	Age_under19_Ptge	Age_19_65_pct	Age_over65_pct	WomanPopulationPtge	ForeignersPtge	SameComAutonPtge	SameComAutonDiffProvPtge	DifComAutonPtge	UnemployLess25_Ptge	Unemploy25_40_Ptge	UnemployMore40_Ptge	AgricultureUnemploymentPtge	IndustryUnemploymentPtge	ConstructionUnemploymentPtge	ServicesUnemploymentPtge	totalEmpresas	Industria	Construccion	ComercTTEHosteleria	Servicios	inmuebles	Pob2010	SUPERFICIE	PobChange_pct	PersonasInmueble	Explotaciones
count	8.119000e+03	8.119000e+03	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8114.000000	7931.000000	7980.000000	8110.000000	8057.000000	7.981000e+03	8.112000e+03	8110.000000	8112.000000	7981.000000	8119.000000
mean	5.741855e+03	4.260666e+03	26.506951	34.403789	48.915409	14.666183	3.019429	13.567747	57.371541	29.073583	47.302755	5.619553	81.629141	4.336688	10.729018	7.322292	37.003976	50.180442	8.400982	10.007836	10.837496	58.649705	398.603032	23.419367	48.878321	146.735265	172.149684	3.246160e+03	5.795812e+03	6214.695257	-4.897406	1.296009	2447.204582
std	4.621520e+04	3.442889e+04	7.540091	16.482285	19.945087	25.093642	2.053726	6.780648	6.818072	11.745849	4.361907	7.348553	12.289063	6.394440	8.847295	9.408555	20.317306	22.803515	12.958405	12.528441	13.281177	24.259562	4219.366083	158.610811	421.863266	1233.023418	2446.812300	2.431471e+04	4.753568e+04	9218.194603	10.383417	0.566620	15062.738051
min	5.000000e+00	5.000000e+00	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	23.459000	0.000000	11.765000	-8.960000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	6.000000e+00	5.000000e+00	2.578400	-52.270000	0.110000	1.000000
25%	1.660000e+02	1.400000e+02	21.678000	21.892500	38.690500	0.759500	1.389000	8.334000	53.845000	19.824500	45.725000	1.060000	75.806000	0.676000	4.933000	0.000000	28.571000	41.667000	0.000000	0.000000	0.000000	50.000000	7.000000	0.000000	0.000000	0.000000	0.000000	1.800000e+02	1.777500e+02	1839.191800	-10.400000	0.850000	22.000000
50%	5.490000e+02	4.470000e+02	26.429000	35.165000	51.582000	1.883000	2.978000	13.889000	58.655000	27.559000	48.485000	3.590000	84.493000	2.190000	8.271000	5.882000	39.935000	50.000000	3.493000	7.143000	8.333000	62.018000	30.000000	0.000000	0.000000	0.000000	0.000000	4.860000e+02	5.820000e+02	3487.737450	-4.960000	1.250000	52.000000
75%	2.427500e+03	1.846500e+03	31.475000	46.032000	62.201000	16.497000	4.533000	19.058500	61.818000	36.908000	50.000000	8.180000	90.462000	5.277000	13.898000	10.469500	46.667000	60.039000	11.732500	14.286000	14.286000	72.123000	147.000000	14.000000	25.000000	65.000000	40.000000	1.589000e+03	2.483000e+03	6893.877800	0.092500	1.730000	137.000000
max	3.141991e+06	2.363829e+06	57.576000	94.117000	100.000000	100.000000	13.245000	33.696000	100.002000	76.471000	72.683000	71.470000	127.156000	67.308000	100.000000	100.000000	100.000000	100.000000	100.000000	100.000000	100.000000	100.000000	299397.000000	10521.000000	30343.000000	80856.000000	177677.000000	1.615548e+06	3.273049e+06	175022.910000	138.460000	3.330000	99999.000000

	Population	AbstentionPtge	Izda_Pct	Age_over65_pct	WomanPopulationPtge	UnemployLess25_Ptge	UnemployMore40_Ptge	SUPERFICIE	totalEmpresas	Pob2010
CCAA
Extremadura	336	20.213	60.444	26.785	44.048	2.381	42.857	4507.5593	15.0	326.0
Extremadura	429	25.275	54.779	30.304	50.117	16.216	51.351	6270.7646	11.0	459.0
Extremadura	569	27.241	44.203	36.028	49.033	8.197	55.738	5702.1000	49.0	674.0
Extremadura	822	30.114	50.813	24.940	51.095	7.407	31.481	9106.4649	50.0	842.0
Extremadura	623	30.185	44.562	25.042	48.154	15.385	36.538	4007.6141	22.0	625.0

	Population	AbstentionPtge	Izda_Pct	Age_over65_pct	WomanPopulationPtge	UnemployLess25_Ptge	UnemployMore40_Ptge	SUPERFICIE	totalEmpresas	Pob2010
count	8.119000e+03	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8119.000000	8110.000000	8114.000000	8.112000e+03
mean	5.741855e+03	26.506951	34.403789	29.073583	47.302755	7.322292	50.180442	6214.695257	398.603032	5.795812e+03
std	4.621520e+04	7.540091	16.482285	11.745849	4.361907	9.408555	22.803515	9218.194603	4219.366083	4.753568e+04
min	5.000000e+00	0.000000	0.000000	0.000000	11.765000	0.000000	0.000000	2.578400	0.000000	5.000000e+00
25%	1.660000e+02	21.678000	21.892500	19.824500	45.725000	0.000000	41.667000	1839.191800	7.000000	1.777500e+02
50%	5.490000e+02	26.429000	35.165000	27.559000	48.485000	5.882000	50.000000	3487.737450	30.000000	5.820000e+02
75%	2.427500e+03	31.475000	46.032000	36.908000	50.000000	10.469500	60.039000	6893.877800	147.000000	2.483000e+03
max	3.141991e+06	57.576000	94.117000	76.471000	72.683000	100.000000	100.000000	175022.910000	299397.000000	3.273049e+06

	Population	AbstentionPtge	Izda_Pct	Age_over65_pct	WomanPopulationPtge	UnemployLess25_Ptge	UnemployMore40_Ptge	SUPERFICIE	totalEmpresas	Pob2010	prop_missings
CCAA
Extremadura	940	31.946	44.970	17.554	47.553	10.714	32.143	47959.3300	NaN	NaN	25.000000
Galicia	1233	40.716	27.345	38.119	51.419	5.556	57.407	NaN	78.0	1361.0	11.111111
Galicia	6691	36.351	17.908	33.881	52.130	4.396	57.418	NaN	363.0	7313.0	11.111111
Galicia	5253	30.209	21.003	33.905	49.476	0.000	0.000	NaN	NaN	NaN	42.857143
Andalucía	783	19.741	76.411	20.945	49.042	20.270	39.189	NaN	NaN	NaN	42.857143
Andalucía	2111	24.924	68.740	17.527	50.734	12.500	40.972	NaN	NaN	NaN	42.857143
Andalucía	3341	38.566	62.968	19.185	51.182	15.734	43.706	23561.5152	138.0	NaN	11.111111
Galicia	3368	26.168	25.262	22.208	50.980	4.153	55.272	NaN	170.0	3469.0	11.111111
Galicia	1828	36.409	22.640	42.997	53.282	2.857	55.714	NaN	71.0	2299.0	11.111111
Cataluña	5839	36.145	28.822	14.010	50.488	9.763	46.450	5430.0184	259.0	NaN	11.111111
Melilla	85584	48.650	34.417	9.684	49.064	16.363	37.472	NaN	4349.0	76034.0	11.111111
Extremadura	2527	31.865	50.908	17.967	50.297	8.040	53.769	NaN	NaN	NaN	42.857143

Ejercicio 5¶

1. Escoge 10 variables numéricas presentes en el conjunto de datos previamente depurado¶

Vamos a trabajar los missing values:¶

2- Agrega los valores por CCAA, distinguiendo entre valores relativos (agregación por la media) y valores absolutos (agregación por suma)¶

3. Valora la necesidad de escalar los datos y decide el tipo de distancia a aplicar¶

4. Explora los métodos de clustering jerárquico para estos datos y decide el tipo de Linkage más adecuado¶

5- Toma una decisión sobre el número de clusters a considerar y realiza un análisis cluster mediante el método k-means.¶

6- Muestra el biplot y comenta los grupos formados. Interpreta los centroides.¶

	Population	SUPERFICIE
CCAA
Andalucía	8387381.0	8.669027e+06
Aragón	1317847.0	4.763550e+06
Asturias	1051229.0	1.027269e+06
Baleares	1104479.0	4.991700e+05
Canarias	2100306.0	7.383646e+05
Cantabria	585179.0	5.246334e+05
CastillaLeón	2472052.0	9.410554e+06
CastillaMancha	2059191.0	7.932435e+06
Cataluña	7502267.0	3.197172e+06
Ceuta	84263.0	1.341337e+03
ComValenciana	4980689.0	2.352726e+06
Extremadura	1088694.0	4.155715e+06
Galicia	2713974.0	2.917790e+06
Madrid	6436996.0	8.064702e+05
Melilla	85584.0	4.024701e+03
Murcia	1467288.0	1.108311e+06
Navarra	640476.0	9.881824e+05
PaísVasco	2189257.0	7.109612e+05
Rioja	317053.0	5.205546e+05

	0	1	2	3	4	5	6	7	8	9
0	2.424811	2.053424	1.812292	2.402924	-0.188271	1.745616	-0.361957	0.660484	1.634669	-1.058098
1	-0.463244	0.721302	-0.455273	-0.459073	-0.675233	0.449608	1.246967	-1.405240	-0.468850	-0.280527
2	-0.572163	-0.553108	-0.582411	-0.566136	0.482354	1.226167	0.791831	0.934803	-0.264290	0.186232
3	-0.550409	-0.733237	-0.459290	-0.557291	0.457386	0.717020	-0.770648	0.716454	-0.183520	0.506710
4	-0.143593	-0.651650	-0.194672	-0.144743	0.625625	0.289376	-0.795087	0.711487	-0.752895	0.689002
5	-0.762554	-0.724552	-0.752781	-0.766647	-0.430350	0.123727	0.059372	-0.281646	-0.058067	0.283558
6	0.008273	2.306352	-0.055561	0.034948	-0.835772	-0.516634	1.690206	-1.479892	-0.429486	0.599982
7	-0.160389	1.802179	-0.250166	-0.152952	-0.984813	0.503426	1.107276	-1.073996	-0.134061	0.039870
8	2.063223	0.187026	2.419259	2.053076	0.551545	-2.608144	-0.312619	0.113611	-0.903846	1.436159
9	-0.967189	-0.903042	-0.945585	-0.975136	2.292182	-0.371739	-1.652666	0.600635	1.927642	-2.516340
10	1.033105	-0.101007	0.990730	1.074881	-1.092321	-1.369092	0.060201	0.432126	-0.521932	0.646128
11	-0.556858	0.513976	-0.596009	-0.556814	-0.479018	1.394964	0.600187	0.524929	1.017650	0.101065
12	0.107104	0.091731	0.149926	0.123893	0.092106	-1.368914	1.119298	1.379020	-0.949855	1.582843
13	1.628037	-0.628420	1.967445	1.623729	-0.670087	0.175102	-1.008045	0.203599	-0.553557	0.559740
14	-0.966650	-0.902127	-0.942250	-0.976988	2.456483	-0.238540	-1.827585	0.489608	2.332989	-2.089914
15	-0.402194	-0.525465	-0.444135	-0.412262	-0.363588	0.041284	-0.937115	0.544308	0.347563	-0.546993
16	-0.739964	-0.566440	-0.717698	-0.748444	-0.013399	0.316104	0.290531	-0.727792	0.104087	-0.138635
17	-0.107254	-0.660997	-0.107691	-0.120369	0.243904	-0.537432	-0.476268	0.359175	-0.679001	-0.365439
18	-0.872090	-0.725943	-0.836131	-0.876596	-1.468731	0.028101	1.176121	-2.701673	-1.465239	0.364658

	Population	SUPERFICIE	totalEmpresas	Pob2010	AbstentionPtge	Izda_Pct	Age_over65_pct	WomanPopulationPtge	UnemployLess25_Ptge	UnemployMore40_Ptge	cluster_kmeans
PC1	-0.307732	-0.306864	-0.301913	-0.310262	0.399468	0.067233	-0.324144	0.181005	0.323635	-0.358863	-0.297107
PC2	0.416770	-0.058824	0.434959	0.415537	0.221417	-0.243070	-0.347585	0.366550	0.109908	0.008755	-0.292477

	PC1	PC2	cluster_kmeans
Ceuta	4.776327	0.958855	0
Melilla	4.864795	1.032096	0
Cataluña	-2.440261	3.830334	1
ComValenciana	-1.529535	1.747195	1
Galicia	-0.928213	0.791379	1
Madrid	-1.417782	2.680517	1
Navarra	0.693211	-1.365749	2
Asturias	0.701941	-0.875903	2
Baleares	1.044479	-0.206041	2
Canarias	0.482153	0.332124	2
Cantabria	0.537125	-1.204489	2
Murcia	1.084965	-0.063255	2
PaísVasco	0.467143	0.271167	2
Extremadura	0.431680	-1.131705	2
Rioja	-1.366352	-3.254432	3
CastillaLeón	-2.564761	-1.689988	3
Aragón	-1.069790	-2.202332	3
CastillaMancha	-1.673014	-1.795331	3
Andalucía	-2.094109	2.145558	4

	Population	SUPERFICIE	totalEmpresas	Pob2010	AbstentionPtge	Izda_Pct	Age_over65_pct	WomanPopulationPtge	UnemployLess25_Ptge	UnemployMore40_Ptge
CCAA
Andalucía	8387381.0	8.669027e+06	489098.0	8370975.0	28.705762	55.122694	20.945208	49.362575	13.890823	42.875152
Aragón	1317847.0	4.763550e+06	90048.0	1347095.0	25.033557	41.598175	33.307423	45.753094	6.444003	46.946937
Asturias	1051229.0	1.027269e+06	67674.0	1084341.0	33.762987	49.701974	29.810372	49.841897	7.168179	49.391141
Baleares	1104479.0	4.991700e+05	89341.0	1106049.0	33.574701	44.388761	17.805015	49.460373	7.454119	51.069343
Canarias	2100306.0	7.383646e+05	135909.0	2118519.0	34.843398	39.926080	17.617239	49.451693	5.438432	52.023920
Cantabria	585179.0	5.246334e+05	37692.0	592250.0	26.880235	38.197441	24.182500	47.716373	7.898245	49.900794
CastillaLeón	2472052.0	9.410554e+06	160390.0	2559515.0	23.822925	31.514937	36.713069	45.622654	6.583358	51.557764
CastillaMancha	2059191.0	7.932435e+06	126143.0	2098373.0	22.698995	42.159800	32.234108	46.331884	7.629214	48.624712
Cataluña	7502267.0	3.197172e+06	595913.0	7512381.0	34.284757	9.688943	21.324291	48.407013	4.904039	55.936445
Ceuta	84263.0	1.341337e+03	3762.0	80579.0	47.411000	33.027000	11.028000	49.258000	14.928000	35.239000
ComValenciana	4980689.0	2.352726e+06	344518.0	5111706.0	21.888273	22.619101	24.188871	48.963561	6.256081	51.799408
Extremadura	1088694.0	4.155715e+06	65281.0	1107220.0	26.513229	51.463457	28.337870	49.125717	11.706470	48.945161
Galicia	2713974.0	2.917790e+06	196552.0	2777804.0	30.820100	22.620951	32.326472	50.618087	4.741159	56.704563
Madrid	6436996.0	8.064702e+05	516402.0	6458684.0	25.072363	38.733570	15.980966	48.564251	6.144123	51.347034
Melilla	85584.0	4.024701e+03	4349.0	76034.0	48.650000	34.417000	9.684000	49.064000	16.363000	37.472000
Murcia	1467288.0	1.108311e+06	92008.0	1461979.0	27.383689	37.337111	16.525956	49.159578	9.334244	45.551578
Navarra	640476.0	9.881824e+05	43866.0	636924.0	30.024482	40.204996	25.958614	46.936813	8.472298	47.689963
PaísVasco	2189257.0	7.109612e+05	151216.0	2178339.0	31.964825	31.297908	20.066888	48.836092	5.700032	46.502295
Rioja	317053.0	5.205546e+05	23024.0	322415.0	19.049741	37.199534	32.763080	43.487810	2.916609	50.325477

	Population	SUPERFICIE	totalEmpresas	Pob2010	AbstentionPtge	Izda_Pct	Age_over65_pct	WomanPopulationPtge	UnemployLess25_Ptge	UnemployMore40_Ptge	cluster_kmeans
CCAA
Ceuta	84263.0	1.341337e+03	3762.0	80579.0	47.411000	33.027000	11.028000	49.258000	14.928000	35.239000	0
Melilla	85584.0	4.024701e+03	4349.0	76034.0	48.650000	34.417000	9.684000	49.064000	16.363000	37.472000	0
Cataluña	7502267.0	3.197172e+06	595913.0	7512381.0	34.284757	9.688943	21.324291	48.407013	4.904039	55.936445	1
ComValenciana	4980689.0	2.352726e+06	344518.0	5111706.0	21.888273	22.619101	24.188871	48.963561	6.256081	51.799408	1
Galicia	2713974.0	2.917790e+06	196552.0	2777804.0	30.820100	22.620951	32.326472	50.618087	4.741159	56.704563	1
Madrid	6436996.0	8.064702e+05	516402.0	6458684.0	25.072363	38.733570	15.980966	48.564251	6.144123	51.347034	1
Navarra	640476.0	9.881824e+05	43866.0	636924.0	30.024482	40.204996	25.958614	46.936813	8.472298	47.689963	2
Asturias	1051229.0	1.027269e+06	67674.0	1084341.0	33.762987	49.701974	29.810372	49.841897	7.168179	49.391141	2
Baleares	1104479.0	4.991700e+05	89341.0	1106049.0	33.574701	44.388761	17.805015	49.460373	7.454119	51.069343	2
Canarias	2100306.0	7.383646e+05	135909.0	2118519.0	34.843398	39.926080	17.617239	49.451693	5.438432	52.023920	2
Cantabria	585179.0	5.246334e+05	37692.0	592250.0	26.880235	38.197441	24.182500	47.716373	7.898245	49.900794	2
Murcia	1467288.0	1.108311e+06	92008.0	1461979.0	27.383689	37.337111	16.525956	49.159578	9.334244	45.551578	2
PaísVasco	2189257.0	7.109612e+05	151216.0	2178339.0	31.964825	31.297908	20.066888	48.836092	5.700032	46.502295	2
Extremadura	1088694.0	4.155715e+06	65281.0	1107220.0	26.513229	51.463457	28.337870	49.125717	11.706470	48.945161	2
Rioja	317053.0	5.205546e+05	23024.0	322415.0	19.049741	37.199534	32.763080	43.487810	2.916609	50.325477	3
CastillaLeón	2472052.0	9.410554e+06	160390.0	2559515.0	23.822925	31.514937	36.713069	45.622654	6.583358	51.557764	3
Aragón	1317847.0	4.763550e+06	90048.0	1347095.0	25.033557	41.598175	33.307423	45.753094	6.444003	46.946937	3
CastillaMancha	2059191.0	7.932435e+06	126143.0	2098373.0	22.698995	42.159800	32.234108	46.331884	7.629214	48.624712	3
Andalucía	8387381.0	8.669027e+06	489098.0	8370975.0	28.705762	55.122694	20.945208	49.362575	13.890823	42.875152	4